The long-term research objectives of this project are: (a) to identify the experimental conditions under which implicit memory for new verbal associations can be developed in normals as well as amnesics, and (b) to specify the cognitive processes that "unitize" or bind the elements of these associations. Two major classes of theories of memory - the systems approach and the processing approach -provide the framework for the proposed studies. In order to understand the conditions and processes that mediate implicit memory for new verbal associations, an experimental analysis of three pivotal issues unexamined in the existing literature is proposed. The first set of studies will examine the role of repetition in binding the elements of the new association at the perceptual level, at the conceptual level, or both. The second set of studies will determine the role of cue constraints in the priming of new verbal associations in amnesia. It is hypothesized here that perceptually constrained cues will be more effective in reducing interference from other responses and in accessing the unitized representations of new associations. The third set of studies will manipulate proactive interference to assess whether new associations in amnesia are perceptual and/or conceptual in nature. Manipulations of these variables will provide converging evidence on the nature of unitization of new associations in implicit memory. Accuracy data in test performance will be collected from three groups of subjects, amnesic patients, matched controls-Implicit, and matched controls- Explicit. These studies are expected to provide mutual knowledge about the functioning of implicit memory in normals as well as amnesics. Specifically, these studies will identify the mechanisms by which new implicit memories may develop in normals when explicit memory contamination is greatly reduced. An understanding of the facilitatory and inhibitory factors in the development of new memories in amnesics has important implications for the management of amnesia.